Effect of ramosetron for preventing postoperative nausea and vomiting: An updated meta-analysis with trial sequential analysis

Ayako Yokoi, Takahiro Mihara, Koui Ka, Takahisa Goto
Department of Anesthesiology, Kanagawa Children’s Medical Center
Department of Anesthesiology and Critical Care Medicine, Yokohama City University Graduate School of Medicine

BACKGROUND
In 2013, we reported a meta-analysis of effect of ramosetron for preventing postoperative nausea and vomiting (PONV). After that, several randomized controlled trials (RCTs) about the effect of ramosetron have been published. When updating meta-analysis, trial sequential analysis (TSA) is recommended to use. TSA adjusts the P value and 95% confidence interval (CI) to prevent the inflation of type I error rate due to viewing the evidence multiple times. The aim of this study was to update the meta-analysis by using TSA.
METHODS
We searched MEDLINE, CENTRAL, Embase, and Web of Science. RCTs that reported the effect of ramosetron for preventing PONV compared with placebo were included. We defined postoperative 0-6 and 6-24 hours as early and late periods, respectively. We collected and analyzed the data of postoperative nausea (PON) and vomiting (POV) separately. We used random-effects model to combine the data. In the TSA, risk of type 1 error was maintained at 5% with a power of 80%, and TSA adjusted P values and CIs were calculated. Heterogeneity was quantified using I2 statistics. Publication bias was evaluated using funnel plot. Sensitivity analysis restricting to double-blind RCTs were conducted.
RESULTS
Of 352 articles initially searched, 20 articles (2003 patients) were included. Conventional random-effects meta-analysis showed that ramosetron was effective for preventing PON and POV in both early and late periods compared to placebo. PON early: risk ratio (RR) [95% confidence interval] = 0.53 [0.41-0.68] (Figure 1A), PON late: 0.47 [0.34-0.66], POV early: 0.49 [0.32-0.74], and POV late: 0.50 [0.36-0.69]. The sensitivity analysis did not change the results. Additionally, the Z curve crossed the TSA monitoring boundary (Figure 1B), which means that the TSA adjusted 95% CI did not include value of 1, in all results.
CONCLUSION
Ramosetron is effective for preventing PONV compared with placebo. TSA suggests that the effect of ramosetron appears to be strong enough to establish a reasonable conclusion and there is no need to conduct RCTs for comparing ramosetron with placebo.

Efficacy of ramosetron in preventing postoperative nausea and vomiting: an updated meta-analysis with trial sequential analysis